Pharmacology of Kv1.7 channels .1 TEA sensitivity

Wild type mKv1.7 channels showed high affinity to TEA with an IC50 of 125µM while the human Kv1.7 channels and the mKv1.7-T0 currents were blocked with an IC50 of about 2.5mM TEA, being about 20fold than the IC50 for wild type channels.

The sequences of mKv1.7 wt and T0 are virtually identical, with the only difference

being the presence of the inactivation domain in the wild type channels. Since it is known that TEA interacts with the ion channel pore the question arises: How is the deletion of the N-terminus part can cause the observed functional effects? It is known that the "voltage dependent gating" of Kv channels is produced by the effects of transmembrane electric potential on the positively charged fourth domain S4. The nature of the ionic species flowing through the permeation pathway is known to influence the binding of the quaternary ammonium compound TEA (Thompson &

Begenesich, 2003). The block of mKv1.7 wt channels by TEA is voltage dependent. It has been suggested that the voltage dependence of the block by TEA of Shaker channels is related to the movement of potassium ions between the two positions of the selectivity filter that are within the membrane electric field. Thus, extracellular TEA can block Shaker B channels deprived from the inactivation domain (D6-46) only when the ion in the selectivity filter is away from the most external site of the ion pathway. It has been extensively documented that discrete amino acids in the extracellular mouth of the ion pore influence dramatically the sensitivity for the block of TEA as well as other blocking agents (MacKinnon & Yellen, 1990; MacKinnon &

Miller, 1989). One such amino acid critical for the interaction of TEA and charybdotoxin with Shaker channel is T449 located in the C-terminal end of the S5-S6 linker. A conservative change in T449 to tyrosine generated a 50fold increase in the sensitivity of the toxin for the channel (from 27 to 0.5mM; MacKinnon & Yellen, 1990). The equivalent position of T449 is occupied by a hydrophobic amino acid in mKv1.7 channels (alanine 398). Another critical amino acid for the binding of charybdotoxin to Shaker channels is in position 425, in the S5-S6 linker, where a mutation of the natural occurring F to an H generated a 8fold increase in the affinity (Perez-Cornejo et al., 1998). Kv1.7 channels naturally contain a histidine in the homologous position (mKv1.7 wt H374, hKv1.7 H341) and therefore it is expected to observe a high affinity to extracellular TEA. Moreover the amino acid residue 373 from mKv1.7 is a serine instead of threonine in the conserved positions 424 and 340 of Shaker and hKv1.7 channels, respectively.

Interestingly, the affinity of mKv1.7 wt for TEA is ~125mM. However, an important influence of TEA on the fast inactivation could only be observed in mKv1.7 wt channels since both the human and the T0 mutants lack this fragment in the NH2 -domain. Accordingly, the affinity of these two channels was 20 times lower for hKv1.7

and mouse mutant both bearing an alanine (IC50 ~2.5mM) at the homologous position of the Shaker binding site for extracellular TEA. A possible explanation of this observation could be that the observed functional effects are due to conformational changes occurring in the intracellular domains that are linked to channel gating. The amino terminal deletion could either change the equilibrium between intrinsic conformational states of the channel protein, or cause a more focalized effect such as electrostatic interactions within the permeation pathway. For the inactivation ball to gain access to the transmembrane pore region through a lateral opening to gain access to the pore it must move through the T1 central cavity, suggesting that the cavity could influence or be influenced by the presence of the inactivating particle (Zhou et al., 2001). One possible interpretation is that the presence of the N-terminus from mKv1.7 wt could favor a high affinity configuration for TEA. As shown in figure 37 there is an apparent voltage dependence of the block of mKv1.7 wt currents by TEA, but considering the model from Thomson and Begenisich it is hard to imagine how a positively charged particle like the inactivation domain of mKv1.7 channels occluding the pore from the intracellular side could induce the proposed high affinity K⁺ ions configuration in the selectivity filter with a K⁺ ion occupying the innermost location in the filter. In the later case a contrary effect would be expected due to steric and/or electrostatic interaction between the inactivation particle from the cavity of the channel, the K⁺ ions on the selectivity filter and the TEA from the pore mouth. However, the report of Thompson and Begenisich is based on the interactions of TEA with Shaker channels that contain a threonine in position 449 (Thompson & Begenisich, 2003). The Kv1.7 channels contain an alanine instead, which is less bulky and hydrophilic than a T. Another possible explanation of the observed differences in the TEA sensitivity could be related to higher affinity of TEA to the N-inactivated mKv1.7 wt channels where the voltage dependence would come from unblock upon the channel opening induced by depolarization.

4.6.2 Zinc block of Kv1.7 currents

The interaction of zinc with Kv1.7 channels resulted in profound block of the current. This block is voltage dependent being about 2fold more sensitive at 0mV than at +40mV for Kv1.7 wt and 3fold different in the mKv1.7-T0 despite that the later

is less sensitive to zinc (mKv1.7 wt IC50 at 0mV =0.23±0.01mM; IC50 at 40mV

=0.43±0.002mM; mKv1.7-T0 IC50 at 0mV =0.67±0.1mM; IC50 at 40mV =1.9±0.6mM).

The voltage dependence of the block by Zn²⁺ of mKv1.7 currents suggests, as the block by TEA, that the blocker has more affinity for predominantly N-type inactivated channel than for the slower inactivating one. The potential site for action of Zn²⁺ block has been suggested to depend on binding of Zn²⁺ to imidazole rings of histidine or to sulphur atoms in cysteines. Lower affinity has been reported to occur at the side chains of acidic amino groups over glutamate or aspartame residues (Vallee & Auld, 1990). It has been reported that a histidine residue located on the channel turret region in position 463 in the human Kv1.5 channels is a high affinity binding site for Zn²⁺. Also residues E456A, D469 and D485 coordinate Zn²⁺ for block of hKv1.5 currents. Kv1.7 channels contain all those residues at the homologous positions (E367, H374, E379 and D396) suggesting that block should occur by a similar mechanism. However, the conformational modification associated to the N-type inactivation of mKv1.7 wt channels must have an influence on the current sensitivity to the block as evidenced by a smaller IC50 at the potentials tested.

4.6.3 kM-conotoxin RIIIK

In this study we showed that kM-conotoxin RIIIK blocks the mouse and human Kv1.7 channels with an affinity of about 150nM and 800nM, respectively. From the other mammalian members of the Kv1 family only Kv1.2 shows affinity on the nanomolar range while all the other tested so far were insensitive to concentrations between 5-10mM of the peptide (Ferber et al., 2004). Interestingly, k-conotoxin PVIIA with pharmacological properties similar to kM-conotoxin RIIIK does not block Kv1.7 channels (IC50> 5 0mM). Affinities for kM-RIIIK were calculated from single concentration experiments where only reduction on the peak currents from single pulses was taken into account. Further experiments with dose response curve determinations have suggested that the block is different for the different conformations of the channel (state dependence). State dependence has been previously shown for the interaction of k-conotoxin PVIIA and kM-conotoxin RIIIK with Shaker channels (Terlau et al., 1999; Ferber et al., 2003) and human Kv1.2 channels (Ferber et al., 2004). Furthermore there seems to be different cooperativity

in the block of kM-RIIIK over Kv1.7 channels (data not shown).

4.6.4 Conkunitzin-S

The mouse and human Kv1.7 channels are the only members of the Kv1 family of mammalian voltage activated potassium channels that are sensitive to Conkunitzin-S with an affinity in the low nanomolar range. Although the IC50 of the mouse channels to the peptide toxin are about 2.5fold higher than the human IC50. Hence, Conkunitzin-S is the only channel blocker that showed a different pattern having the highest affinity for the human channel instead of the mouse mKv1.7 wild type channel. Further studies on the kinetics parameters of the block of Conkunitzin-S might contribute to the clarification of the factors determining the higher affinity that the other blockers assayed had to the fast inactivating channel Kv1.7 from mouse in this work. The specificity of Conkunitzin-S indicated that this peptide might be a good tool for studying Kv1.7 channels in vitro and in vivo.

4.6.5 Oxidation sensitivity of Kv1.7 channels

It is known that fast inactivating Shaker B channels exposed to oxidizing agents recover from inactivation faster than reduced ones. The underlying mechanism has been suggested to be a direct enzymatic oxidation of the third methionine in the N-terminal inactivation domain of the channel that destabilizes the properties of the inactivation particle eliminating fast inactivation (Ciorba et al., 1997, 1999). Kv1.4, a mammalian channel present in cardiac tissue, is also regulated by oxidation causing slower inactivating currents which are due to patch excision or oxidation of cysteine13 on the N-terminus, resulting in loss of fast inactivation (Ruppersberg et al., 1991). Several other examples of modulation of potassium currents by oxidation have been documented in the literature (Duprat et al., 1995;

Szabo et al., 1997, Vega-Saenz de Miera & Rudy, 1992). Accordingly, oxidation of key amino acids in the pore domain of Shaker channels (lacking the N-inactivation domain) has been implicated in the acceleration of P-type inactivation in those channels (Chen et al., 2000). We have observed that mKv1.7 wt channel inactivates

significantly faster upon patch excision and in presence of the oxidizing agent dithiodipyridine while washing lead to reestablishment of slower inactivation kinetics.

The oxidative modulation of the Kv1.7 current does not only affect the inactivation time constant but causes a decrease in the current amplitude. We could not observe a faster inactivation of mKv1.7-T0 currents upon patch excision nor speeding of the current decay in the presence of DTDP in the human Kv1.7 channels. That indicates the relevance of the N-terminal domain in the expression of the phenomenon. The analysis of the N-terminus of mKv1.7 wt channels revealed the presence of several features that strongly suggest its involvement as an inactivation particle with characteristics closely resembling the ball domain from Shaker. More over, the investigation of the structural aspects of the inactivation domain of Kv1.7 revealed the presence of a particular pattern of an iron-sulfur cluster (FeS) coordinated by a cysteine cluster motif (CxxCxxxCx27H) similar to the one that was described for the nqo3 subunit of the proton translocating NADH-Quinone oxidoreductase (NDH-1) from Thermus thermophilus HB-8 (Nakamura-Ogiso et al., 2002). Iron-sulfur proteins are ubiquitous in nature containing sites with one to eight iron atoms, sometimes repeated within the same protein (Beinert et al., 1997). Iron-sulfur clusters have a remarkable facility for conversion and interconversion in the free and protein bound conditions supporting the concept that they are modular structures that can undergo ligand exchange reactions and oxidative degradation, both of high biological significance. Enzymes bearing FeS clusters like aconitase are modulated by changes in the pH or thiol attack (Kennedy et al., 1984). From the structure/function relation point of view it is particularly interesting the finding of an iron-sulfur coordination motif in the N-terminus of mKv1.7 wt channels. Such motives are known to bind cysteine ligands from different subunits where iron sulfur clusters effect dimer formation as in the Fe protein nitrogenase (Howard & Rees, 1996). Even more exciting is the fact that iron-sulfur clusters serve as sensors of iron, dioxigen, superoxide ion (O2-) and possibly nitric oxide (Rouault et al., 1992; Hentze & Kühn, 1996; Beinert & Kiley, 1996; Gaudu & Weiss, 1996; Hidalgo et al., 1995). There have been at least two modes of sensing described so far (for a review of the mechanisms see Beinert et al., 1997). Therefore, supported on the results that neither mKv1.7-T0 nor the human Kv1.7 showed sensitivity to redox modulation, it seems plausible that the oxidation sensitive region (or an important component of it) of the Kv1.7 channel must lie in the amino terminal part of the protein. A similar arrangement HX5CX20CC has been

suggested to coordinate Zn²⁺ in the N-terminal, cytoplasmic tetramerization domain (T1) of Shab, Shaw and Shal subfamily members, but is not found in Shaker subfamily members (Bixbi et al., 1999). The presence of a motif used by many other proteins for sensing the redox state of the cell might indicate that at least for mKv1.7 wt channels the formation of an iron-sulfur cluster in the inactivation domain or between the N-terminus of different subunits could function as the redox sensor that triggers different responses including closure in response to determined stimulus.

Furthermore, such biochemical structures are known to drive rearrangement of the protein structure by pure electrochemical coordination as a mean for fast and sensitive mechanism of functioning.